Accent Normalization Using Self-Supervised Discrete Tokens with Non-Parallel Data

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Bai, Qibing, Inoue, Sho, Wang, Shuai, Jiang, Zhongjie, Wang, Yannan, Li, Haizhou
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911073223835648
author Bai, Qibing
Inoue, Sho
Wang, Shuai
Jiang, Zhongjie
Wang, Yannan
Li, Haizhou
author_facet Bai, Qibing
Inoue, Sho
Wang, Shuai
Jiang, Zhongjie
Wang, Yannan
Li, Haizhou
contents Accent normalization converts foreign-accented speech into native-like speech while preserving speaker identity. We propose a novel pipeline using self-supervised discrete tokens and non-parallel training data. The system extracts tokens from source speech, converts them through a dedicated model, and synthesizes the output using flow matching. Our method demonstrates superior performance over a frame-to-frame baseline in naturalness, accentedness reduction, and timbre preservation across multiple English accents. Through token-level phonetic analysis, we validate the effectiveness of our token-based approach. We also develop two duration preservation methods, suitable for applications such as dubbing.
format Preprint
id arxiv_https___arxiv_org_abs_2507_17735
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Accent Normalization Using Self-Supervised Discrete Tokens with Non-Parallel Data
Bai, Qibing
Inoue, Sho
Wang, Shuai
Jiang, Zhongjie
Wang, Yannan
Li, Haizhou
Audio and Speech Processing
Sound
Accent normalization converts foreign-accented speech into native-like speech while preserving speaker identity. We propose a novel pipeline using self-supervised discrete tokens and non-parallel training data. The system extracts tokens from source speech, converts them through a dedicated model, and synthesizes the output using flow matching. Our method demonstrates superior performance over a frame-to-frame baseline in naturalness, accentedness reduction, and timbre preservation across multiple English accents. Through token-level phonetic analysis, we validate the effectiveness of our token-based approach. We also develop two duration preservation methods, suitable for applications such as dubbing.
title Accent Normalization Using Self-Supervised Discrete Tokens with Non-Parallel Data
topic Audio and Speech Processing
Sound
url https://arxiv.org/abs/2507.17735